John E. Kuhn

Fractures of the acromion process: A proposed classification system

Summary: A review of 27 fractures of the acromion process during a 15-year period revealed five distinct types that were classified into three groups. Stress fractures are rare, do not result from acute trauma, and gain little benefit from nonoperative treatment. Type I fractures are minimally displaced. Type IA fractures are avulsion fractures and heal rapidly. Type IB fractures result from direct trauma to the extremity, and are minimally displaced. Most heal with nonoperative treatment. Type II fractures are displaced laterally, superiorly or anteriorly and do not reduce the subacromial space. Most are pain free with full motion after 6 weeks of nonoperative treatment. Type III fractures reduce the subacromial space. This may occur by an inferiorly displaced acromion fracture, or an acromion fracture associated with an ipsilateral, superiorly displaced glenoid neck fracture. Patients in this group sustained significant trauma to the involved extremity. All type III fractures treated nonoperatively develop significant limited shoulder motion with pain, suggesting that early surgical intervention may be indicated.

The effect of glenoid inclination on superior humeral head migration.

Superior humeral head migration may contribute to the development of rotator cuff disease. The purpose of this study was to test the hypothesis that superior inclination of the glenoid facilitates superior humeral head migration. Eight cadaveric shoulders were tested by use of a custom test fixture, and rotator cuff forces were applied. Glenoid inclination was varied (intact, +5 degrees, +10 degrees, and +15 degrees ), and the force required to produce superior humeral head migration was measured. Each increase in glenoid inclination (more superiorly facing glenoid) produced significant reduction in the force required for superior humeral head migration (5 degrees, 14.2% reduction; 10 degrees, 29.9% reduction; and 15 degrees, 37.5% reduction; P <.001). These findings demonstrate that glenoid inclination is an important factor in determining the force required for superior humeral head migration. This suggests that a more upward-facing glenoid increases the risk for superior humeral translation and, in turn, may play a role in the development of rotator cuff disease.

Ligamentous Restraints to External Rotation of the Humerus in the Late-Cocking Phase of Throwing: A Cadaveric Biomechanical Investigation

The late-cocking phase of throwing is characterized by extreme external rotation of the abducted arm; repeated stress in this position is a potential source of glenohumeral joint laxity. To determine the ligamentous restraints for external rotation in this position, 20 cadaver shoulders (mean age, 65 16 years) were dissected, leaving the rotator cuff tendons, coracoacromial ligament, glenohumeral capsule and ligaments, and coracohumeral ligament intact. The combined superior and middle glenohumeral ligaments, anterior band of the inferior glenohumeral ligament, and the entire inferior glenohumeral ligament were marked with sutures during arthroscopy. Specimens were mounted in a testing apparatus to simulate the late-cocking position. Forces of 22 N were applied to each of the rotator cuff tendons. An external rotation torque (0.06 N m/sec to a peak of 3.4 N m) was applied to the humerus of each specimen with the capsule intact and again after a single randomly chosen ligament was cut (N 5 in each group). Cutting the entire inferior glenohumeral ligament resulted in the greatest increase in external rotation (10.2° 4.9°). This was not significantly different from sectioning the coracohumeral ligament (8.6° 7.3°). The anterior band of the inferior glenohumeral ligament (2.7° 1.5°) and the superior and middle glenohumeral ligaments (0.7° 0.3°) were significantly less important in limiting external rotation.

Infusion of five percent dextrose increases mortality and morbidity following six minutes of cardiac arrest in resuscitated dogs

UDDEN CARDIAC ARREST is a leading cause of mortality in the United States, accounting for approximately 384,000 deaths per year.’ The standardization of cardiac life support protocols, the establishment of rapid response emergency medical systems, and the education and certification of both medical and lay personnel have reduced the mortality associated with sudden cardiac arrest; however, mortality rates in most areas are still approximately 85%.2 Cardiac arrest protocols include the initial establishment of a venous access to facilitate drug administration. The patency of this access is maintained with a continuous infusion of fluids which routinely contain 5% dextrose. When an individual is found unconscious and no history is available, an ampule of 50% dextrose is frequently administered to correct for possible hypoglycemia. These uses of dextrose are generally considered either beneficial or at least innocuous. The treatment groups in this study were specifically designed to evaluate the use of dextrose in an experimental model mimicking the clinical setting of cardiac arrest.

Arthroscopic Capsular Release for the Stiff Shoulder Description of Technique and Anatomic Considerations

The anatomic proximity of several neurovascular structures remains a major concern to the surgeon interested in performing arthroscopic capsular re lease. We evaluated the anatomic relationships be tween the released capsule and the axillary nerve, posterior circumflex humeral artery, and brachial ar tery in a frozen cadaveric model. With the aid of electrocautery, seven cadaveric shoulders under went complete arthroscopic capsular release. The release was performed circumferentially, approxi mately 1 cm lateral to the glenoid rim. All shoulders were subsequently frozen and sectioned through the plane of the capsular release while the shoulder was maintained in the lateral arthroscopic position (45° of abduction and 20° of flexion). Anatomic dissection revealed an average distance from the capsular re lease to the axillary nerve of 7.04 mm (95% confi dence interval, 5.62, 8.47), to the posterior circum flex humeral artery of 8.2 mm (95% confidence interval, 6.41, 9.99), and to the brachial artery of 15.97 mm (95% confidence interval, 9.85, 22.09). As the axillary nerve was followed medially from the released capsule, the inferior border of the subscap ularis muscle became interposed between the cap sule and the axillary nerve. This limited anatomic study shows that a relatively safe margin between the capsule and the neighboring neurovascular structures can be obtained by releasing the capsule within 1 cm of the glenoid rim.

Ibuprofen improves survival and neurologic outcome after resuscitation from cardiac arrest

Post-ischemic inflammatory changes in the central nervous system (CNS) following cardiac arrest and resuscitation are potentially responsible for ultimate survival and much of the neurologic damage, producing greater morbidity and mortality in successfully resuscitated patients. This study was undertaken to assess the non-steroidal anti-inflammatory agent, ibuprofen, in a controlled and monitored experimental model of canine cardiac arrest and resuscitation. With the investigator blinded as to the intervention, eight of 21 dogs were randomly assigned to receive ibuprofen as an i.v. bolus (10 mg/kg) and a 6-h i.v. infusion (5 mg/kg per h). The other 13 dogs received an equivalent volume of 0.9% NaCl to serve as controls. No statistically significant differences between the two groups were detected in any pre-arrest variables. All 21 dogs were successfully resuscitated. At 24 h, dogs receiving ibuprofen exhibited 100% survival, while control dogs exhibited only 54% survival (P = 0.03). The majority of deaths for the control group occurred within the first 6 h. Neurologic deficit scores were assigned at 1, 2, 6 and 24 h after resuscitation. A general trend occurred such that dogs treated with ibuprofen improved over time, while the control dogs remained severely impaired. A significant difference in neurologic deficit score was detected at 6 h (P = 0.01). At 24 h the ibuprofen group exhibited minimal neurologic deficit (5.9 +/- 3.2), and the control group exhibited significantly more severe neurologic impairment (52.2 +/- 13.0, P = 0.01). These results suggest that ibuprofen may be helpful in the pharmacologic management of cardiac arrest as a means of increasing survival and decreasing neurologic impairment.

A Statistics Primer Validity and Reliability

Measurement is a fundamental activity of everyday life. We weigh ourselves on the bathroom scale, estimate how much gas we have left before the next exit on the interstate, pour 2 tablespoons of cough syrup for a seasonal cold, estimate how far we can hit a baseball, and note the growth of our children in pencil on the kitchen wall. The importance of such measures vary with circumstances and persons. For example, measuring cough syrup for an adult probably does not require the same degree of precision as measuring such a drug for an infant. Reliance on the accuracy of the gas gauge while driving in a major city is perhaps not as crucial as when driving in Death Valley. Likewise, measurement is a fundamental and essential component of orthopaedic research. We measure range of motion, Cobb angles, anterior knee laxity, angulation of humeral fractures, and so forth. And from these measurements we make diagnoses, for example, clinically significant progression of spinal curvature or unstable knees. When measurements are reported as part of clinical research, readers should be able to evaluate how well the investigators made these measurements. Observations and evidence gathered to answer a study question must be reliable and accurate so that statistical analyses based on these measurements are relevant. That is, before statistical tests are conducted on the study data, there should be some assessment of the reliability and validity of the study measurements. Generally, a discussion of the source, measurement procedures, reliability, and relevance is found in the “Materials and Methods” section of an article reporting research results. Unfortunately, discussions regarding issues of measurements in clinical research are frequently absent. Readers should exercise great caution when considering conclusions from published results of a study that have not included an account of how well the observations and data have been measured. Specifically, the reader wants to know about the quality of the data. Have the authors successfully measured what they have purported to measure, and have they made these measures in such a way as to be reproducible? Validity, as applied to data measurement, refers to the degree to which the measurement represents a true value. How accurately does a single measure of blood pressure represent the patient’s true blood pressure? How accurately does an arthrometer estimate knee joint laxity? Reliability, on the other hand, refers to the ability of the researchers to reproduce or repeat the same measurements. Assuming that the patient’s blood pressure does not change, how closely is the same researcher able to reproduce the same measure of blood pressure? Are other researchers able to obtain similar blood pressure readings? How reliably is the physician able to measure knee laxity using an arthrometer? Is he or she able to obtain the same value more than once? Will other physicians obtain the same result? Validity and reliability can be related (and perhaps intertwined) because errors in a study can be caused by either. Several examples will illustrate these concepts. Consider a situation in which five medical students measure blood pressure on the same patient. The values they obtain range from 105/72 to 145/96 mm Hg. It is clear that these measurements lack reliability. Such differences in blood pressure may be due to the subject’s inability to relax among so many “doctors-in-training,” lack of hearing acuity among the students, or cuff inflation variability. Now suppose that these students were able to recruit a more relaxed subject, to reduce extraneous noise, and to acquire a digital read-out manometer. Then, suppose that in a similar exercise all five students obtained identical blood pressures of 145/90 mm Hg. However, it is later * Address correspondence and reprint requests to Mary Lou V. H. Greenfield, MPH, MS, University of Michigan, Orthopaedic Surgery, TC2914G-0328, 1500 East Medical Center Drive, Ann Arbor, MI 48109. No author or related institution has received any financial benefit from research in this study. 0363-5465/98/2626-0483

A Statistics Primer

02.00/0 THE AMERICAN JOURNAL OF SPORTS MEDICINE, Vol. 26, No. 3

A Statistics Primer Power Analysis and Sample Size Determination

on potential clients, consumers need data to compare products, and voters require information on economic trends to help with decision-making at election time. Sports fans everywhere are masters of data consumption: batting averages, passing efficiency, shooting percentages, and other facts. Not surprisingly, the scientific community requires even more data. Unfortunately, data are only useful when the masses of numbers that we collect can be put into usable order. Fortunately, this ordering of masses of numbers can be accomplished by an understanding of statistics.

ASSESSMENT OF OUTCOME IN SHOULDER ARTHROPLASTY

groups. From a review of the literature and from their own clinical experience, the investigators anticipate approximately 50% of the women with a regular training program will suffer an ACL tear. They hypothesize that a reduction by one-half, resulting in only a 25% incidence in ACL injuries in the hamstrings training group, would be a clinically significant finding. The investigators set their